VMware Cloud Community
PappaMangino
Contributor
Contributor

Network Capture Sample for Converter P2V Traffic

      Hi

Yesterday I reported a problem with conversions failing on moving to the second disk on a target machine.  The error that I was getting was a block copy failure.  (I also got File copy failures when I increased the block size.)    My transfer rate on all of these test was an abismal 600-700kbs

I did some more testing and found that this issue seems to be related to the transfer rates of the current task.

What I was doing (because my test lab is remote from the location of the hosts) was to locate a local test ESX host that had enough capacity to test this box.   (Capacity is why we are using the remote test ESX host)  and that succeeded.    My transfer rate was about 5.7mbs

I ran a second test to the original host but using a local server - transfer rate was better than 700k but still was only about 1.5mbs.  Not as good as the previous test.

I did a network capture while the job seemed to be running fine and then I set up a capture at the point I predicted the job woud fail (~68% based on disk size).

The job failed when switching to the second drive.

I have my captures - does anyone have a sample capture file or an explanation of the expected traffic during a successful transfer that I can use as a guide.   What would be really nice is if someone has developed a parser for Netmon with this traffic. 

Thanks.

GM

0 Kudos
5 Replies
DSTAVERT
Immortal
Immortal

Not sure what your intended purpose is here. Could you provide a clearer picture of what you are trying to accomplish with Converter.

-- David -- VMware Communities Moderator
0 Kudos
PappaMangino
Contributor
Contributor

Hi David

This is really for my own education.   I am trying to troubleshoot what appears to be a Network issue and I would like to look at some standard data so that I can do some modelling (what is the threshold when this will cause the converter to fail and include in my documentation for my customer.

Thanks.

0 Kudos
DSTAVERT
Immortal
Immortal

The problem is there isn't going to be any standard data. Every environment will be different and provide different numbers. You mention local and remote. Where are you cloning from and to??

See whether these help sorting out some of your issues http://kb.vmware.com/kb/1004087 http://kb.vmware.com/kb/1006607

-- David -- VMware Communities Moderator
0 Kudos
PappaMangino
Contributor
Contributor

Thanks - reviewing the KB articles now.   What I am really looking for is what does the typical conversation look like -

There are my observations -

I am copying 4K Blocks for BLock Level copy

When I look at the CAP I see vmware-convereter.exe start SSL Communications -

16k of data is being transferred followed by 4 or 5 ACKs from the ESX host.

My problem is coming from a specific location (Pitts to Cleve) the transfer rate is like 500-650 kbs (which is awful) and when the rate is this low the cloning of the second drive (from c: to e:) always fails like seconds into the copy.   I have been told that this pipe should not be experiencing this type of transfer. 

I will look at the articles - hoping they can shed some light on this.

Thanks

0 Kudos
DSTAVERT
Immortal
Immortal

I assume Pitts to Cleve is Pittsburgh and Cleveland. I don't know how fat your connection is but I would consider exporting the VM as an OVA and then transmitting the file. An OVA is highly compressed and is easily imported using the vSphere Client.

-- David -- VMware Communities Moderator
0 Kudos